9250
J. Phys. Chem. B 2008, 112, 9250–9257
Peptide Cysteine Thiyl Radicals Abstract Hydrogen Atoms from Surrounding Amino Acids: The Photolysis of a Cystine Containing Model Peptide Olivier Mozziconacci,† Victor Sharov,† Todd D. Williams,‡ Bruce A. Kerwin,# and Christian Scho¨neich*,† The Department of Pharmaceutical Chemistry, 2095 Constant AVenue, UniVersity of Kansas, Lawrence, Kansas 66047, The Mass Spectrometry Laboratory, UniVersity of Kansas, Lawrence, Kansas 66045, and Department of Process and Product DeVelopment, Amgen Inc., 1201 Amgen Court West, Seattle, WA ReceiVed: February 28, 2008; ReVised Manuscript ReceiVed: April 30, 2008
Peptide cysteine thiyl radicals were generated through UV-photolysis of disulfide precursors, in order to follow intramolecular reactions of those radicals with neighboring amino acids. When reactions were carried out in D2O, there was a significant incorporation of deuterium specifically into the CR-H bonds of glycine residues in positions i+1 and i-1 to the Cys residue, indicating a fast reversible H-atom transfer. This H-atom transfer occurred prior to the formation of final, nonradical products including free thiol, thioaldehyde, and aldehyde. Such fast H-atom transfer is relevant to biologic conditions of oxidative stress and to the stabilization of proteins against oxidation, where the formation of carbon-centered radicals in proteins may lead to fragmentation, intramolecular cross-linking, aggregation and/or epimerization. 1. Introduction CysS•,
Cysteine thiyl radicals, represent important intermediates during catalytic turnover of several enzymes such as, e.g., the ribonucleotide reductases.1 CysS• are formed during oneelectron oxidation of Cys,2–5 or one-electron reduction of disulfide bridges.6,7 Specifically, the photoionization of protein Trp (or Tyr) residues generates electrons, which subsequently reduce nearby protein disulfide bonds, as exemplified in a series of mechanistic studies on cutinase,8,9 bovine growth hormone,10 and R-lactalbumin.11,12 In R-lactalbumin, CysS• are proposed to ultimately recombine with Trp-derived radicals to generate intraprotein cross-links. However, both experimental and theoretical data suggest that, within a protein, CysS• may be involved in a series of additional reactions with its surrounding amino acids before final recombination. For example, theoretical calculations13,14 predict that within random peptide and β-sheet conformations but not the R-helix,15 the CR-H bond dissociation energy (BDE) of Gly is lower compared to the BDE of the S-H bond, resulting in a thermodynamic preference for hydrogen atom abstraction from Gly by CysS• (Scheme 1, reaction 1). In fact, our own experimental data reveal second order rate constants on the order of 103-105 M-1s-1 for hydrogen abstraction by thiyl radicals from the CR-H bond of Val, Ala, and Gly,16,17 and from the side chains of Phe, Met, Ser, and Thr.16 Moreover, some evidence for the addition of CysS• to the aromatic ring of Phe was provided,18 supported by theoretical calculations (S. Naumov, C. Scho¨neich, unpublished results). Considering the important role of CysS• during enzyme turnover, as well as potentially in the biologically relevant formation of S-nitroso-cysteine from Cys and nitric oxide (NO),19 a chemical evaluation of CysS• reactivity toward surrounding amino acids in peptides and proteins is warranted. Such information is also of immense value for the biotechnology industry for the * To whom correspondence should be addressed. FAX: (785) 864-5736. E-mail:
[email protected]. † Department of Pharmaceutical Chemistry, University of Kansas. ‡ Mass Spectrometry Laboratory. # Department of Process and Product Development.
following reasons. Currently, a large effort is placed on the production of antibodies for diagnostic and therapeutic purposes. Antibodies contain multiple disulfide bonds, and, similar to many other proteins, are sensitive to light-induced degradation.20 Light-induced fragmentation of the disulfide bond according to a mechanism similar to that detailed for R-lactalbumin will generate antibody thiyl radicals with all possible consequences for intraprotein hydrogen abstraction reactions. The latter may lead to antibody aggregation and fragmentation, two phenomena, which are frequently observed during antibody production and formulation studies.21 Ultimately, these mechanisms may provide the chemical basis for increased immunogenicity or loss of activity of the antibody products. In the present study, we have generated thiyl radicals of a disulfide-linked model peptide, (GlyGlyCysGlyGlyLeu)2 1 (Scheme 2), through 253.7 nm photolysis of the disulfide bridge specifically designed to follow thiyl radical reaction in peptides. We report on the mechanisms of product formation as well as covalent H/D-exchange of peptide C-H bonds when the reaction is carried out in D2O, induced through reversible H-atom transfer to a peptide thiyl radical, analyzed by liquid chromatography and nanoelectrospray ionization (LC/NSI-MS) spectrometry, followed by MS/MS analysis. A probabilistic model rationalizes specifically the regioselectivity for H-atom abstraction. 2. Experimental Section 2.1. Materials and Reactions. The peptide (structure 1) was provided by Amgen Inc. (Thousand Oaks, CA) at a purity level of >95% and used without further purification. The peptide was dissolved in H2O (MilliporeQ) or in deuterium oxide (Cambridge Isotope Laboratories, Inc., 99.9%) at a concentration of 100 µM. Prior to UV-irradiation a 200 µL aliquot was placed in a quartz tube. The samples were gently saturated with Ar or O2 for 5 min prior to photolysis. The samples were irradiated for up to 10 min with four UV lamps (Southern England, RMA500) at 253.7 nm in a Rayonet (Southern England, U.S.A.) photochemical reactor.
10.1021/jp801753d CCC: $40.75 2008 American Chemical Society Published on Web 07/09/2008
Thiyl Radical and Cysteine Containing Model Peptide
J. Phys. Chem. B, Vol. 112, No. 30, 2008 9251
SCHEME 1: Reaction Scheme for Covalent H/D Exchangea
a The intramolecular H-transfer occurs between the thiyl radical group of the cysteine residue and CR-H. The mechanism was confirmed by the incorporation of a deuterium atom in the final product.
SCHEME 2: Formation of the Main Products Obtained after UV-Irradiation of Ar-Saturated Aqueous Solution Containing Peptide 1
2.2. Sample Preparation for Capillary HPLC-MS Analysis. Immediately after photoirradiation, the samples were diluted 10-fold in MilliporeQ water. The samples were injected onto a capillary column (C18, 5 cm × 150 µm, Micro-Tech Scientific, CA) and eluted with a linear gradient delivered at rate of 10 µL min-1. Mobile phases consisted of water/methanol/formic acid at a ratio of 95, 5, 0.01% (v:v:v) for solvent A and a ratio of 20, 80, 0.01% (v:v:v) for solvent B. The following gradient was set: 95% of solvent A for 2 min, followed by an increase of solvent B from 5 to 80% within 12 min. 2.3. Nano-Electrospray Ionization Time-of-Flight (ESI TOF MS) Analysis. ESI-MS spectra were acquired on a Q-TOF-2 (Micromass Ltd., Manchester,U.K.) hybrid mass spectrometer operated in the MS1 mode and acquiring data with the time-of-flight analyzer. The instrument was operated for maximum resolution with all lenses optimized on the [M + 2H]2+ ion from the cyclic peptide Gramicidin S. The cone voltage was 35 eV and Ar was admitted to the collision cell at a pressure that attenuates the beam to about 20% and the cell
was operated at 12 eV (maximum transmission). Spectra were acquired at 16 129 Hz pusher frequency covering the mass range 350-2000 amu (amu ) atomic mass unit) and accumulating data for 3 s per cycle. Time-to-mass calibration was made with CsI cluster ions acquired under the same conditions. 2.4. Quantification of Free Thiols. Thiols were quantified by reaction with 5,5′-dithiobis(2-nitrobenzoate) (DTNB; Ellman’s reagent).22 Titration with 200 µM DTNB was carried out in phosphate buffer (100 mM, pH 7.5) and the absorption of the solution (λmax ) 412 nm, ε ) 14 100 M-1 cm-1) measured with a Cary 50 Bio spectrophotometer (Varian, Palo Alto, CA), calibrated with known concentrations of reduced glutathione between 0-100 µM. 2.5. MS/MS Analysis. CID spectra were acquired by setting the MS1 quadrupole to transmit a precursor mass window of ( 1.5 amu centered on the most abundant isotopomer. Ar was the collision gas admitted at a density that attenuates the beam to 20%; this corresponds to 16 psi on the supply regulator or 5.3 × 10-5 mbar on a penning gauge near the collision cell.
9252 J. Phys. Chem. B, Vol. 112, No. 30, 2008 CHART 1: Representation of the Virtual Projection of the Cr Atoms of Product 3 along an Infinite Line (xi Values)a
a
The b-type fragments are represented by the labels b1, b2, b3, b4,
b5.
The collision energy varied between 20-35 eV. Spectra were acquired for 2-3 min in 5 s cycles as the peptides were eluted off a desalting column. 2.6. Quantum Yields. The chemical actinometer KI/KIO3 was used to calculate the fluence (UV dose) of our photoreactor according to a published procedure.23 On exposure to UV light (253.7 nm), the KI/KIO3 system forms triiodide, which was determined spectrophotochemically and used to determine the UV fluence (I0) using a quantum yield (QY) at 254 nm of 0.73.23 I0 was equal to 2.96 × 10-6 quanta/min. 2.7. Covalent H/D Exchange and Isotopic Correction. The deuterium composition of peptide ions and their fragments was determined from the differences between the average mass of a covalently deuterated peptide and the average mass of the corresponding fully protonated peptide. The average masses were calculated from centroided isotopic distributions. The distribution of deuterium incorporation was obtained after isotopic correction by subtracting the isotope abundance distribution in the product formed during UV-irradiation in H2O from the isotope abundance distribution of the same product generated in D2O. 2.8. The Law of Probability. Information on the sequence of any peptide provided by MS/MS fragmentation can be compared to a cumulative mathematical step-function, F(x) ) ∑in) 1 Pi(X ) xi), where X is a discrete random variable taking the values xi with the probability Pi. All these parameters were set to quantify the density of probability of exchange of one hydrogen by one deuteron in a linear peptide containing one CysS•. The variable X was defined as an event which leads to the incorporation of one deuteron within the linear peptide product GGCGGL (containing a reduced Cys residue) in the probabilistic space, restricted to the internal space of the analyzer (Q-TOF). The discrete xi values were arbitrarily chosen as the virtual projections of the CR position of each amino acid on a virtual infinite line (Chart 1). Topologically, a peptide can be seen as a continuous electron distribution from the N- to the C-termini. Consequently, the discrete values xi were fitted using a continuous mathematical cumulative function: the integrated function Φ(xi) ) ∫x-∞ fµ,σ(x)dx, where fµ,σ represents the Gaussian function 1/σ2πe-1/2(x-µ/σ)2, and µ and σ account for the average and the standard deviation, respectively. During the fragmentation of GGCGGL starting at the N-terminus and ending at the C-terminus, Φ(-3) and Φ(+3) were set to 0 and 1, respectively. According to the labeling of the CR positions (Chart 1), in the range of [-∞;-3], Φ(-3), the space is free of peptide. Thus, the probability distribution to detect the incorporation of one deuteron within the peptide is zero. Conversely, if the probability distribution is applied to the interval [-∞;+3], Φ(+3), the entire peptide is covered. Therefore, if the peptide has incorporated one deuteron, the probability to detect that incorporation within
Mozziconacci et al. the peptide fragments is equal to 1. These initial parameters are used to normalize the deuterium incorporation into the peptide and its fragments. 3. Results 3.1. Structural Analysis of the Products. The identification of products generated through UV-irradiation of peptide 1 was carried out by HPLC-MS followed by MS/MS analysis. Irradiations were carried out in Ar-, O2-, and air-saturated solutions. The reaction pathways leading to the formation the products will be developed in the discussion section. 3.1.1. Ar-Saturated Solutions. HPLC-MS analysis highlighted two main products with m/z 445.2 and 463.2, eluted at 3.4 and 7.4 min, respectively (Figure 1A). These products were identified as peptides 3 and 5 (Scheme 2) by MS/MS analysis (Figure S1 and S2, Supporting Information). A scan of all peaks between 3 and 13 min (Figure 1, B) reveals additional reaction products with m/z 461.2 and 427.2, respectively, assigned to the peptides 4 and 6, confirmed by MS/MS analysis (Scheme 2, and Figure S1, Supporting Information). A product with m/z 889.4 coeluted with the starting peptide 1 (Figure 1B, inset) and was assigned by MS/MS to structure 7 (Scheme 2), which represents a Schiff base formed by condensation of aldehyde 5 with the N-terminal amino group of product 3 (Scheme 2, Figure S2, bottom panel, Supporting Information). 3.1.2. O2-Saturated Solutions. The main products 3 and 5 (m/z ) 463.2 and 445.2) detected after UV-irradiation in Arsaturated solution were also detected in O2-saturated samples (Figure 1C,D). Two of the intermediates identified in Arsaturated solution, 4 and 6 (m/z ) 461.2; 427.2), could be separated from the main products (Figure 1C). A new product with m/z 511.2 formed only in the presence of O2, eluted at 3.8 min, and was identified through MS/MS analysis as product 8 (Scheme 3, Figure S3, Supporting Information). 3.1.3. Air-Saturated Solutions. The products described for Ar- and O2-saturated solutions were also observed in airsaturated solutions. Two minor products with m/z 792.3 and 810.4 were observed in air-saturated solutions after taking into account the whole time window of the HPLC-MS scan (Figure 1E). These products were identified by MS/MS analysis (Figure S4, Supporting Information) and assigned to the structures 9 and 10, respectively (Scheme 4). 3.1.4. Quantification of the Main Products. The photolytic yields of the major products were quantified representatively for air-saturated solutions. The degradation of peptide 1 under UV-irradiation was followed by HPLC separation and linear regression of the peak area versus concentrations (Figure 2A, 9). The formation of product 3 was followed by DTNB titration (Figure 2A, b). The moles of peptide 1 consumed and the moles of product 3 formed under irradiation were utilized to calibrate peak intensities for 1 and 3 in the mass spectrometry scans. The relative quantitative data for the other photoproducts were obtained by comparison of their peak intensities to those of 1 and 3, assuming similar response factors. The yields of all photoproducts obtained after 10 min of irradiation are summarized in Table 1. 3.2. Covalent H/D Exchange in the Cysteine-Containing Product 3. Photoirradiation of 1 was carried out in D2O or H2O after saturation of the solutions with either Ar, O2, or a mixture of Ar and O2 (nine independent measurements) and the isotopic correction performed for product 3 as described in the Experimental Section. Product 3 was selected for this type of analysis as it is an abundant product and directly results from thiyl radical 2. Therefore, it is expected to contain all H/D-exchanged sites originating from reaction of the thiyl radical prior to dispro-
Thiyl Radical and Cysteine Containing Model Peptide
J. Phys. Chem. B, Vol. 112, No. 30, 2008 9253
Figure 1. HPLC-MS traces and mass spectra (A) HPLC trace monitored by total ion current obtained through UV-irradiation of an Ar-saturated solution of peptide 1. (B) Mass spectrometric scan of all products eluting between 3 and 13 min during HPLC analysis as displayed in panel A. Inset: mass spectrometric scan of all products eluting between 13.5 and 14.5 min. (C) HPLC trace obtained after UV-irradiation of an O2-saturated aqueous solution of peptide 1. (D) Mass spectrometric scan of all products eluting between 2 and 13 min during HPLC analysis as displayed in panel C. (E) Mass spectrometric scan obtained after UV-irradiation of peptide 1 in air-saturated solution.
SCHEME 3: Formation of Specific Products Obtained after UV-Irradiation of O2-Saturated Aqueous Solution Containing Peptide 1
portionation (Scheme 2, reaction 5). The incorporation of a single D will be indicated through an increase of the mass spectrometric signal with m/z 464.2, which is the monoisotopic peak of product 3, where one H-atom has been replaced by one D-atom. MS analysis revealed a 63% increase of the base peak intensity (BPI) relative to the signal of the nondeuterated peptide with m/z 463.2 (Figure 3A). No H/D-exchange was observed
in nonirradiated control samples. Isotopic correction was applied to the MS/MS results of 3 in order to estimate the relative increase of the BPI of the b2, b3, b4, and b5 fragments in order to quantitate H/D-exchange at the specific amino acid residues after photoirradiation in D2O versus H2O. The low mass of fragment b1 was outside the range of the mass analyses. For application of the law of probability, the BPI values of the
9254 J. Phys. Chem. B, Vol. 112, No. 30, 2008
Mozziconacci et al.
SCHEME 4: Formation of the Minor Products Obtained after UV-Irradiation of Air-Saturated Aqueous Solution Containing Peptide 1
b-fragments were normalized by assigning the value of the initial parameter Φ(+3) ) 1 to the BPI of (m/z +1) of the pattern ion (463.2 +1) (Figure 3C, full circles). After isotopic correction
TABLE 1: Yields of the Main Products Generated after 10 Min of UV-Irradiation of Peptide 1 in Air-Saturated H2O [M+H]+ of products
structure
concentration (µM)
463.2 445.2 427.2 461.2 511.2
3 5 6 4 8
35 22 7
